for t, batch in target_iter: x, y, _ = batch if self.cuda: x = x.cuda() y = y.cuda() features = self.feature_extractor.forward(x) task_out = self.task_classifier.forward(features) class_output = F.softmax(task_out, dim=1) pred_task = class_output.data.max(1, keepdim=True)[1] n_correct += pred_task.eq(y.data.view_as(pred_task)).cpu().sum() n_total += x.size(0) acc = n_correct.item() * 1.0 / n_total return acc if __name__ == '__main__': dataset = 'LABELME' data_path = './vlcs/'+dataset+'/test/' img_transform = transforms.Compose([transforms.Resize((225, 225)), transforms.ToTensor(), transforms.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225])]) dataset = Loader_validation(path=data_path, transform=img_transform) loader = torch.utils.data.DataLoader(dataset=dataset, batch_size=64, shuffle=True, num_workers=4) cp_path = './' cp_name = '_checkpoint_9ep' cuda = True test_object = OffTest(cp_path, cp_name, loader, cuda) test_object.load_checkpoint() acc = test_object.test() print(acc)
source_loader = torch.utils.data.DataLoader(dataset=source_dataset,
batch_size=args.batch_size,
shuffle=True,
num_workers=args.workers)
test_source_dataset = Loader_unif_sampling(hdf_path1=test_source_1,
hdf_path2=test_source_2,
hdf_path3=test_source_3,
transform=img_transform_test)
test_source_loader = torch.utils.data.DataLoader(
dataset=test_source_dataset,
batch_size=args.batch_size,
shuffle=True,
num_workers=args.workers)
target_dataset = Loader_validation(hdf_path=target_path,
transform=img_transform_test)
target_loader = torch.utils.data.DataLoader(dataset=target_dataset,
batch_size=args.batch_size,
shuffle=True,
num_workers=args.workers)
task_classifier = models.task_classifier()
domain_discriminator_list = []
for i in range(3):
if args.rp_size == 4096:
disc = models.domain_discriminator_ablation_RP(
optim.SGD, args.lr_domain, args.momentum_domain,
args.l2).train()
else:
disc = models.domain_discriminator(args.rp_size, optim.SGD,
args.lr_domain,
# Setting seed random.seed(seeds[run]) torch.manual_seed(seeds[run]) checkpoint_path = os.path.join(args.checkpoint_path, args.target+'_seed'+str(seeds[run])) if args.cuda: torch.cuda.manual_seed(args.seed) img_transform_train = transforms.Compose([transforms.RandomResizedCrop(225, scale=(0.7,1.0)), transforms.RandomHorizontalFlip(), transforms.ToTensor(), transforms.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225])]) img_transform_test = transforms.Compose([transforms.Resize((225, 225)), transforms.ToTensor(), transforms.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225])]) train_source_1 = args.data_path + 'train_' + args.source1 + '.hdf' test_source_1 = args.data_path + 'test_' + args.source1 + '.hdf' source_dataset = Loader_validation(hdf_path=train_source_1, transform=img_transform_train) source_loader = torch.utils.data.DataLoader(dataset=source_dataset, batch_size=args.batch_size, shuffle=True, num_workers=args.workers) test_source_dataset = Loader_validation(hdf_path=test_source_1, transform=img_transform_test) test_source_loader = torch.utils.data.DataLoader(dataset=test_source_dataset, batch_size=args.batch_size, shuffle=True, num_workers=args.workers) model = models.AlexNet(num_classes = 7, baseline = True) state_dict = torch.load(args.model_path+'alexnet_caffe.pth.tar') del state_dict["classifier.fc8.weight"] del state_dict["classifier.fc8.bias"] not_loaded = model.load_state_dict(state_dict, strict = False) optimizer = optim.SGD(list(model.classifier.parameters()), lr=args.lr, momentum=args.momentum, weight_decay=args.l2, nesterov=True) if args.cuda: model = model.cuda()